Investing in Talent

The U.S. Talent Pipeline is broken but fixable

Today’s guest post fits with our ethos of democratizing aristocratic education. We build AI tutors like Virgil to make great and difficult works, accessible, meaningful and relevant. If you have a non-obvious take on Education, design, philosophy, and/or technology, we’d love to feature you. Pitch us at zohar@lightningstudios.ai - Zohar

---

Investing in Talent

By Ilana Walder-Biesanz

The year is 1957. The Soviet Union has successfully launched Sputnik 1, and the United States is in a panic. America’s prestige and world leadership – not to mention our unrivaled military capabilities – are at stake.

So we spring into action. We double down on research and space exploration with the creation of DARPA and NASA. We greatly expand the NSF budget. And we pass the National Defense Education Act to modernize curricula, raise standards, and fund selective talent programs to cultivate our best and brightest minds.

Our post-Sputnik focus on talent pipelines was wise. Top talent disproportionately expands the boundaries of our knowledge and capabilities. The top 1% of scientists generate over 20% of citations. The top 1% of inventors are 5-10X more productive than the average inventor. Nowhere is this dynamic clearer than in AI, where firms like Meta are dangling 8- and 9-figure pay packages in front of a small roster of superstar researchers.

Furthermore, we can spot people with high potential for extraordinary contributions at a young age. Of course, no one can say which 8-year-old will go on to win a Nobel Prize, but early signs heavily skew the probabilities of future outcomes. Students who score in the top 5% on third-grade math exams go on to account for 25% of patent filings. Data from the Study of Mathematically Precocious Youth indicate that there seems to be no plateau in the correlation between early math achievement and odds of many indicators of career contributions (e.g., PhD attainment, STEM research publications, patents, high incomes).

–

Fast forward to today. We again have a rival for scientific leadership, with China snapping at America’s heels in biotech, AI research, nuclear fusion, and more.

But since the 1950s, we’ve systematically destroyed much of our national talent infrastructure. The NSF redirected funding away from top talent development; most of the selective summer programs that opened thanks to post-Sputnik funding have shuttered. K12 education legislation (e.g., No Child Left Behind, Every Student Succeeds Act) has tied school accountability to the share of students meeting a minimum level of proficiency, moving focus away from supporting high achievers. Between 2012 and 2018, 1 in 4 schools eliminated gifted programs.

We tend to assume that children who are doing well on tests will turn out fine – that resources are better directed to students who are struggling. But when classes have wide variation in students’ ability levels (as they inevitably do), students at both ends of the curve are poorly served by instruction that caters to the middle. Of course we should help students who lag behind. That produces a fully literate society. We should also challenge and engage those who are capable of racing ahead. That boosts national competitiveness and economic growth.

Cultivating talent starts with early identification. At National Math Stars, we work with schools, districts, and state Departments of Education to look at their existing data sources (e.g., state achievement tests, MAP Growth) and identify top-scoring 2nd- and 3rd-graders. Many schools have never analyzed the data this way before! They’ve searched for kids who need intervention to catch up, but not for kids who are ready for more. Looking at universally administered test scores also improves equity. Yes, rich kids do better than poor kids on tests, but the gap is even larger for qualitative talent identification approaches like teacher nominations.

Existing data provide a start, but top talent identification requires more. Universally administered tests cap out. They can give you a strong sense of who is ready for a gifted program or advanced math, but won’t tell you who is at a 99.9th-percentile level rather than 99th. Since National Math Stars explicitly searches for (and builds our Stars Program for) 1-in-1000 young mathematicians, we conduct additional, higher-ceiling testing to gauge nominees’ numerical, logical, and spatial reasoning skills with greater precision.

Once we’ve found talented students, two key interventions can help nurture their talents: acceleration and enrichment.

Meta-analyses find average learning gains of 0.50-0.70 standard deviations from acceleration, making it one of the most impactful learning interventions in existence. However, philosophical opposition to tracking and acceleration has eliminated this option for advanced students in many areas. This has essentially created a two-tier system: knowledgeable, higher-income parents accelerate their kids through paid providers of additional instruction (e.g., Kumon, Art of Problem Solving, Russian School of Math), while high-achieving children of lower-income or lower-social-capital parents are limited to learning with their age peers.

Acceleration is not the goal per se; preparing students to work at the frontiers of STEM is. For that, enrichment is equally important. National Math Stars ensures top students have access to STEM clubs, competitions, summer camps, supplies, and mentoring, in addition to at-level coursework. A high pre-college dose of STEM opportunities increases the chance of STEM career accomplishments, as does early access to mentorship.

–

I don’t claim this problem is easy to solve at scale. There’s a lot of infrastructure to (re)build: acceleration pathways, gifted programs, selective summer camps, widespread math contest participation, and more. This will take political willpower and significant funding. (Supporting the highest-need, highest-talent students is particularly expensive: National Math Stars currently spends over $20K per year on our fully funded Stars.)

The potential reward is commensurately large. Simply closing income- and demographic-driven gaps in invention rates among the mathematically talented would quadruple U.S. innovation. Our funders have estimated that the net present value of a full package of talent supports, including acceleration for top-10% students, enrichment for top-1% students, and intensive long-term talent development for top-0.1% students, would be in the tens of trillions of dollars.

And maintaining U.S. world leadership? Priceless. But not free.

–

Ilana Walder-Biesanz is the Founder and CEO of National Math Stars, a nonprofit organization that ensures mathematically extraordinary students from all communities have the resources they need to reach the frontiers of math and science.

If you’re interested in learning more about National Math Stars or in collaborating with us as a STEM enrichment partner, funder, state Department of Education, nominating school/district, or in any other capacity, please reach out to ilana@nationalmathstars.org.